Salivary amylase provides an excellent example to illustrate the principles that have emerged from the study of salivary function in the oral cavity, namely, the multifunctionality, amphifunctionality, and the conformational dictum of salivary components. The multifunctional nature of amylase includes: 1) hydrolysis of complex carbohydrates; 2) binding to hydroxyapatite (HAP); and 3) binding to bacteria (e,g, streptococci) in solution and when bound to HAP. The amphifunctionality arises from the final outcome of the functions. In other words, the different functions of a single molecule may be beneficial or potentially harmful depending upon the site of action. For salivary amylase, its binding to bacteria in solution may result in bacterial clearance (protective) while its presence in the enamel pellicle may facilitate dental plaque formation at the enamel surface (harmful). Its binding to viridans streptococci both in solution as well as when bound to the HAP surface is dictated by the maintenance of its native conformation. Alpha- amylases consist of three structural domains, A, B and C. Based on the structural homology among alpha-amylases, a hypothesis that the bacterial and substrate binding functions are associated with distinct, conformationally independent structural domains is proposed. As a corollary to this hypothesis, we propose that the C-domain of amylase is responsible for bacterial binding. As part of a long range goal for the design of amylase molecules for use in artificial salivas, this application will obtain critical baseline information using biophysical and biochemical techniques. Protein X-ray crystallography will be used to obtain the 3D structure of amylase. Biophysical and biochemical techniques will be used to define the structural domains associated with the three functions of amylase listed above and to test whether or not the intact C-domain is solely responsible for bacterial binding. Comparison of the biological properties of native amylase and fragments obtained by selective fragmentation or partial denaturation will be used to identify the structural domains that are responsible for its functions. On the basis of the information derived from these studies, molecular modeling and dynamics will be used to differentiate between the substrate and bacterial binding domains and to identify residues suitable in these domains for future mutagenesis experiments.